Tandem mass spectrometry (MS/MS) for the analysis of drugs and drug metabolites in racing animals

MISSING IMAGE

Material Information

Title:
Tandem mass spectrometry (MS/MS) for the analysis of drugs and drug metabolites in racing animals
Physical Description:
viii, 113 leaves : ill. ; 28 cm.
Language:
English
Creator:
Brotherton, Harry O'Neil, 1950-
Publication Date:

Subjects

Subjects / Keywords:
Drugs -- Analysis   ( lcsh )
Drugs -- Metabolism   ( lcsh )
Mass spectrometry   ( lcsh )
Doping in sports   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1982.
Bibliography:
Includes bibliographical references (leaves 110-112).
Statement of Responsibility:
by Harry O'Neil Brotherton, Jr.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000365760
notis - ACA4577
oclc - 09882309
System ID:
AA00003854:00001

Full Text
















TANDEM MASS SPECTROMETRY (MS/MS) FOR THE
ANALYSIS OF DRUGS AND DRUG METABOLITES
IN RACING ANIMALS












By

HARRY O'NEIL BROTHERTON, JR.


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA
1982















ACKNOWLEDGMENTS

I would like to express my gratitude to Dr. Richard

A. Yost who has directed my research here at the University

of Florida. I would also like to acknowledge the assistance

and guidance of Drs. John Dorsey and J. D. Winefordner.

I am grateful to Dr. Ronald Gronwall and Walter Stone

for providing the equine blood serum which was used in this

study. I would also like to thank Dr. Steve Sundloff for

his assistance in acquiring canine serum and urine samples.

Members of my research group have provided support

and friendship during my studies. I would like to extend

a special word of thanks to Dean Fetterolf and Bob Perchalski

for helpful discussions and assistance.















TABLE OF CONTENTS

Page
ACKNOWLEDGMENTS ---------------------------- ii

ABSTRACT ---------------------------------- yv

CHAPTER

1. INTRODUCTION ------------------------ 1

Background ------------------------ 1
Current Screening Methodology ------ 2
Metabolism and Pharmacokinetics
Studies --------------------------- 5
Tandem Mass Spectrometry ---------- 8
Overview of Research --------------- 10

2. EXPERIMENTAL ------------------------- 14

Instrumentation -------------------- 15
Reagents --------------------------- 17
Sample Preparation ----------------- 17
Deproteination ------------------ -17
Serum Extraction ----------------- 17
Urine Extraction ----------------- 20
Animal Dosage and Sample Collection 20

3. RESULTS AND DISCUSSION --------------- 21

Sample Preparation ----------------- 23
Screening ------------------------- 31
Quantitation ----------------------- 65
Metabolite and Pharmacokinetic
Studies --------------------------- 71
Serum ---------------------------- 72
Urine -------------------------- --78
Pharmacokinetics ----------------- 88

4. CONCLUSIONS AND FUTURE WORK ---------- 93


iii









APPENDIX ---------------------------------------- 95

DRUG DAUGHTER SPECTRA ---------------------- 95

REFERENCES ------------------------------------ 110

BIOGRAPHICAL SKETCH ------------------------------ 113















Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

TANDEM MASS SPECTROMETRY (MS/MS) FOR THE
ANALYSIS OF DRUGS AND DRUG METABOLITES
IN RACING ANIMALS

By

Harry O'Neil Brotherton, Jr.

December, 1982

Chairman: Richard A. Yost
Major Department: Chemistry

Screening for drugs and metabolites in racing animals

is presently a lengthy procedure involving extensive sample

cleanup followed by chromatographic analysis. All positives

are confirmed by a more selective analytical technique,

usually gas chromatography-mass spectrometry (GC/MS). A

rapid, sensitive technique based on tandem mass spectrometry

(MS/MS) has been developed which eliminates the need for

chromatographic analysis as well as greatly reducing the

amount of sample preparation required. The MS/MS technique

replaces the chromatographic separation with a mass

spectrometric separation and allows simultaneous analysis

for 25 to 50 drugs, drug classes, functional groups or

metabolites. The Triple Quadrupole Mass Spectrometer
consists of, in series, a dual chemical ionization/electron









impact (CI/EI) sample ionization source, a quadrupole mass

filter, an RF-only quadrupole, a second quadrupole mass

filter, and an electron multiplier.

A procedure which allows drug screening and confirmation

to be performed more rapidly than current methods has been

developed. The direct screening of blood serum allows rapid

identification at the part per million level but quantitation

appears impractical due to the protein binding of most drugs

and the thermal decomposition of the proteins in the serum

at high temperatures. A simple acid-base extraction of the

serum eliminates this problem and allows the detection and

quantitation of drugs and metabolites at the part per billion

level.

Metabolic analyses were carried out on serum and urine

samples from two greyhounds. The parent spectrum operating

mode of the MS/MS system was used to determine the masses

of the ions present in the physiological fluids which had

the same structural subunits as the parent drug. The

daughter spectra of these probable metabolites were then

examined to determine their structure. Serum samples

collected over six hours were analyzed and pharmacokinetic

data calculated.















CHAPTER 1
INTRODUCTION

Background

The detection, identification and quantitation of

drugs and their metabolites in the blood serum and urine

of racing animals is a challenging analytical problem.

Snall doses of any of a wide variety of drugs, including

stimulants, depressants, narcotics and anesthetics can

be used to alter the performance of racing animals.

Screening for these compounds, as well as studies of metabolic

disorders and pathways, requires the development of effective

testing procedures that are capable of detecting and properly

identifying a large number of drugs and metabolites at

concentrations less than a part per million (yg/mL) in

blood and urine.

The doping of racing animals (the administration of

a substance which can affect the performance of an animal

at the time of racing) can serve any of several pruposes,

one of which is to enable that animal to win. This may be

done by administering a stimulant, such as caffeine or

amphetamine, in a single dose before the race, or by administer-

ing a drug such as a vitamin or hormone, continuously over

a period of time in order to build up a better or stronger





2



animal. A sedative or injurious substance that will impair

the animal's performance may also be administered with the

intent of causing the animal to lose the race. Accidental

administration of a substance which would normally be

classified as a drug, such as its inclusion in the animals'

feed, may also occur. Stopping this illegal alteration in

the performance of the racing animals is the primary objective

of the pre-race and post-race screening procedures currently

in use (1).

Current Screening Methodology

Blood serum and urine are the biological fluids primarily

used in drug testing in North America (2). Most drugs and

metabolites can be detected in both serum and urine, some

more easily in one than the other. Blood is used in both

pre-race and post-race testing while urine is primarily used

in post-race testing. In pre-race testing, blood must be

collected from all horses in every race within two hours of

racing and screened at trackside labs so that doped animals

can be detected. Post-race, urine is collected from selected

animals, frozen, and shipped to off-track laboratories for

analysis. The best drug coverage can be attained by using a

combination of pre-race and post-race testing of serum and

urine, but an affordable alternative is post-race testing

of serum and urine from selected animals.









The analysis of drugs and metabolites in biological

fluids is hindered by their low concentrations and by

interfering substances. The use of preliminary extraction

steps to separate desired compounds or classes of compounds

from the interfering matrix is important in current procedures;

it also allows concentration of the sample. Typical

extraction procedures for a serum sample calls for acid,

neutral and base fractions to be extracted from a 5 mL serum

sample. This may be proceeded by protein denaturing or use

of some type of precipitating agent to free drugs and

metabolites from binding to the protein in the serum. Urine

samples, usually 10 mL, are first hydrolyzed using either

enzymatic or acid hydrolysis and then extracted into acid,

neutral and base fraction (3-5).

Preliminary screening has traditionally been carried out

by thin layer chromatography (TLC) (6). Ten centimeter by

5 to 20 centimeter TLC plates (depending on the number of

samples to be screened) coated with a 0.25 millimeter silica

gel layer are used. The extract fractions are evaporated to

dryness and approximately 50 pL of ethyl acetate added. The

TLC plates are then spotted with approximately 10 pL of the

extract and developed using one of a variety of standard

solvent mixtures. After the solvent front has traveled 5

centimeters, the TLC plates are dried and carefully sprayed

with a series of reagents to visualize any drugs which are

present (5,7).









The TLC plates are first impregnated with a zinc compound

which fluoresces yellow under short wavelength (254 nm) ultra-

violet (UV) light. Compounds which absorb UV light at about

254 nm quench the fluorescence on the plate and appear as

dark spots while some compounds fluoresce themselves and

appear as a spot of another color. The plate is then exposed

to long wavelength UV radiation (350 nm). Compounds which

fluoresce under 350 nm light appear as bright colored spots.

Fluorescamine, a reagent which selectively reacts with

compounds that contain a primary amine, is then sprayed on

the TLC plate. After standing for 5 minutes in a hood, 350

nm UV light is used to identify any new bright yellow spots.

Phenothiazine spray, which reacts with phenothiazines and

their metabolites, is then used and pink and purple spots

observed. Alkaloids and many other types of bases are then

reacted with Dragendorff's Reagent to produce red spots on

the yellow background under 254 nm UV light. Mandelin's

Reagent, which reacts with a variety of compounds to give

spots of different colors, is then used (5,7,8). This

sequence of reagents is used to screen for a large number of

drugs and metabolites, but additional specialized tests are

necessary for a number of compounds, such as despropionyl

fentanyl, and considerable effort is required to develop

tests which can confidently be used to screen for many newly

developed drugs.





5



Each spot observed during the visualization procedure

is marked and the distance the drug traveled relative to

the distance traveled by the solvent front (Rf) is noted.

If a spot on the TLC plate has an Rf and visualization

appearance which corresponds to that of a known drug (a

"suspicious sample"), a new TLC plate is prepared and an

application of sample is placed across several centimeters

of the plate'sorigin. After the plate is developed, the

spot is scraped from the TLC plate, washed with methanol

and the sample is prepared for confirmation through the

use of a more selective analytical technique '(8).

A pharmacological response is produced by most organic

bases at a serum concentration of 1 to 100 ppb and by most

acids and neturals at 0.1 to 50 ppm (9). By concentrating

the sample extract, TLC screening can detect most drugs at

the low part per million level. Additional sensitivity is

obtained by derivitization and use of special visualization

reagents. This produces adequate sensitivity for the

screening of most drugs, but due to the time required for

sample cleanup, derivitization and screening, most analyses

are carried out post-race. Typical turnaround times for

these samples are a month or more, with an average cost of

20 dollars per sample (2).

While TLC is still the most widely used screening procedure

(1,8), developmental work is continuing in an effort to

develop more sensitive, selective and rapid methods of analysis.








Procedures for the use of gas chromatography (GC) (10-12)

and high performance liquid chromatography (HPLC) (13) as

the chromatographic screening step in drug and metabolite

analysis are under development. Reduced sample cleanup

together with automated readout and data evaluation systems

have encouraged their development. The main drawback to

these systems is the necessity of derivitizing a large number

of drugs so that they can be separated chromatographically,

or detected by the detectors in the systems.

Confirmation and/or identification of the suspicious

samples is carried out using more selective analytical

technique, usually gas chromatography/mass spectrometry

(GC/MS) (1)'. A number of other techniques have been used

with varying degrees of success. These include UV spectroscopy,

radioimmuno assay (RIA), infrared (IR) spectroscopy, GC and

HPLC (7). Extensive sample cleanup requirements, interference

and low sensitivities have discouraged the use of most of

these techniques in the confirmation of the presence of illegal

substances in the biological fluids of racing animals.

Metabolism and Pharmacokinetic Studies

The primary function of drug metabolism in the body is

to change a compound structurally so that it becomes less

toxic and is more easily discharged from the body. This is

effected by biochemically transforming drugs from lipid

soluble non-polar compounds to water soluble polar compounds

that are less pharmacologically active and can be more easily

excreted from the body (3:).









The structural determination of metabolites by GC/MS,

the current method of choice (14,15), is a lengthy process.

The method generally involves preliminary sample cleanup

followed by extractions into acid, base and neutral fractions

using procedures discussed above. Derivitization may then

be necessary to increase the volatility, thermal stability

or to change chromatographic characteristics of the metabolites.

Typically a month or more is required for the development

of chromatographic procedures and identification of the

metabolites of a given compound.

Studies of the rate of change of drug concentration in

blood and urine (pharmacokinetics) provides information on

the rates of absorption, distribution, metabolism and excret-

ion of the drug and its metabolites. Pharmacokinetic studies

aid in the determination of dosage schedules for drugs in

man and animals, and can be used to determine the route (oral

or intramuscular) and form (salt or complex) of the drug

administered. These studies are carried out by administering

a single dose of drug (orally, intraveinously or intramuscul-

arly), and collecting blood and/or urine at selected intervals

over a given time period. These samples are then analyzed as

discussed above and the pharmacokinetic characteristics

determined. A considerable variation in pharmacokinetic

activity may exist between any two individuals. These dif-

ferences may be due to genetic differences, differences in

gastrointestinal tract absorption or urinary pH, or the

existence of some disease state in the liver or renal function

(9,16).









Tandem Mass Spectrometry

Mass spectrometry (MS) initially came into analytical

use in the mid 1950's. The first instruments were magnetic

sector instruments and were used primarily for quantitative

analysis of petroleum products. An electric sector was

added to form a double focusing instrument which increased

the available mass resolution by focusing the ions of varying

kinetic energy into the magnetic sector. Quadrupole mass

filters were developed and became commercially available in

the early 1960's. The quadrupole mass filter consists of

four parallel rods with opposing pairs electrically connected.

When RF potentials of opposite phase are applied to the two

pairs of rods, all ions have stable trajectories and pass

through the mass filter. If, in addition, a DC potential

of opposite sign is applied to the two pairs of rods, only

ions that fall in a given m/z range have stable trajectories

and pass through the mass filter. A quadrupole mass spec-

trometer is constructed by placing a source of ions at one

end of the mass filter and an ion detection device at the

other.

Mixture analysis with early mass spectrometers was

made difficult by impurities in the systems producing

mixed mass spectra. The MS system was coupled with a GC

for separation of the mixture components prior to the mass

spectral analysis (17). The GC/MS interface was perfected

and GC/MS became a well established technique for the identi-

fication of trace components in samples that are amenable to

GC analysis.








Since the early 1960's, attempts to develop a practical,

efficient liquid chromatograph/mass spectrometry (LC/MS)

interface have been undertaken by a number of chemists (18).

This technique is particularly attractive to the analytical

chemist because it can more easily separate polar, thermally

labile compounds which must be derivitized if they are to

be analyzed by GC/MS. A number of devices using a variety

of interfacing methods are currently commercially available,

but most of these are impractical or unsatisfactory for

routine LC/MS analysis. The most recent development in

this area, a direct liquid introduction LC interface which

feeds the effluent from a micro LC column directly into the

ion volume of the mass spectrometer (19), appears to be a

promising development in the use of LC/MS.

The need for some type of separation prior to the final

mass analysis in the analysis of complex mixtures by MS is

well established. The preliminary separation should remove

most of the interfering substances and allow the collection

of mass spectra which can be used confidently to identify and

quantitate trace constituents of the mixture. When the

separation prior to mass analysis is carried out by chroma-

tographic separation, the time required to achieve separation

is usually a few minutes to an hour or more. It is, therefore,

desirable to develop some type of preliminary separation which

is more rapid and flexible than a chromatographic method.









Tandem mass spectrometry (MS/MS) is a relatively new

analytical technique which uses a separation step carried

out by a mass analyzer in lieu of a chromatographic

separation, followed by a second step of mass analysis. A

collision chamber is placed between the two mass analyzers.

It can be pressurized with an inert gas in order to fragment

the ions as they pass between the mass analyzers (collisionally

activated dissociation (CAD)), or the ions can be allowed

to fragment unimolecularly (metastable ions). The use of

CAD in conjunction with tandem mass spectrometry has proven

to be very useful for the trace analysis of selected compounds

in complex mixtures (20-24). Pioneering research on reversed-

geometry double-focusing mass spectrometers in the laboratories

of Cooks (23 (Glish et al.), 25 (Kondat & Cooks)) and

McLafferty (26 (McLafferty & Bockhoff)) demonstrated the

potential for the use of MS/MS for mixture analysis.

The subsequent development of the triple quadrupole mass

spectrometer by Yost and Enke (27) has made MS/MS a viable

analytical technique for routine analysis. The triple quad-

rupole system is much simpler than reversed-geometry double-

focusing instruments. Two quadrupole mass filters are used

for mass analysis with a third RF-only quadrupole placed

between the two mass analyzers. This RF-only quadrupole has

no mass filtering action, but is used to focus the ions during

fragmentation. In the reversed-geometry double-focusing

instruments, ions enter the collision chamber with 3 to 30 keV

of kinetic energy,while in the triple quadrupole mass









spectrometer, ions entering the collision chamber have

only 3 to 50 eV of kinetic energy. This energy difference

leads to a fundamentally different fragmentation process

in each system (28), the low kinetic energy fragmentation

process being better suited for analytical applications

since the resulting fragments provide significantly more

useful structural information.

The operating modes of the tandem mass spectrometer (27)

used in the current study are neutral loss scan, parent

scan, daughter scan and selected reaction monitoring (SRM).

During the neutral loss scan, quadrupole one and quadrupole

three scan simultaneously with a fixed mass difference so

that any ion which fragments to lose a selected neutral

fragment can be detected. A parent scan requires quadrupole

one to scan while quadrupole three is set to a fixed mass,

so that any ion which fragments to produce a fragment of

the selected mass is detected. A daughter scan is collected

by passing the parent ion of the compound of interest through

the first quadrupole, fragmentating it by CAD in the second

(RF-only) quadrupole, and then scanning the third quadrupole

to determine all the fragments of the parent ion. During

selected reaction monitoring, only one daughter ion (usually

the most intense) from each parent ion is monitored with

the third quadrupole.









Overview of Research

During the first part of this study, a new MS/MS

procedure for the rapid screening and confirmation of

trace amounts of drugs and metabolites in physiological

fluids was developed. The procedure involves the use of

SRM, neutral loss, or parent scans for preliminary screening

to indicate the possible presence of selected drugs, drug

classes or functional groups, followed by the acquisition of

complete daughter spectra for confirmation. The screening

is carried out with blood serum, urine, or the acid-neutral

and base extracts from the serum or urine. Data are collected

in intervals of 0.1 second or less over the single mass unit

selected for each compound enabling sequential screening for

25 to 50 compounds or drug classes during the approximately

100 seconds that the solids probe is heated from 30 to 325 C.

Then presence of compounds for which the selected reaction

yields signal above background during the screening procedure

is then confirmed by collecting complete daughter spectra of

the indicated compounds' parent ions from a second sample,

and matching them with a library of daughter spectra of pure

compounds.

The structural determination of the metabolites of three

drugs was also carried out using the method described by

Perchalski and Yost (29). The major fragments (structural

subunits) of each drug were determined by collecting a CI









mass spectrum of the pure drug followed by the collection

of a daughter spectrum of each major fragment. All parent

ions from a serum or urine extract which produced one or

more of these fragments were identified using the parent

spectrum operating mode of the tandem mass spectrometer (27).

A full daughter spectrum of each of these parent ions was

then collected and the structure of each parent evaluated to

determine if it was indeed a metabolite of the original drug.

Data collection and evaluation for the metabolites of a drug

can be carried out in less than one day.

Pharamacokinetic data were also collected for two drugs,

diethylcarbamazine and phenylbutazone. Serum samples were

collected from two dogs at hourly intervals for six hours

after oral dosing. These samples were analyzed using the SRM

procedures discussed above.















CHAPTER 2
EXPERIMENTAL

The first objective of these studies was to establish a

set of conditions which could be used for the collection of

the CAD spectra of large numbers of drugs with widely vary-

ing structures. The conditions used had to be sufficiently

gentle so as not to dissociate totally the parent ion of

easily fragmented substances such as methocarbamol, but hard

enough to fragment efficiently the parent ion of drugs which

are difficult to fragment such as morphine. This was done

by collecting daughter spectra of several drugs over a range

of collision energies from 10 to 30 eV and of collision

gas pressures from 1.0 to 2.5 millitorr. As a compromise,

optimum CAD conditions of 18 eV and 1.8 millitorr were

selected for the collision energy and collision gas pressure,

respectively. A library of the CAD daughter spectra was

then constructed to be used in the identification of unknown

drugs which were detected in serum or urine (see Appendix A).

Procedures that were to be used in the analysis for drugs

and metabolites in complex mixtures (blood serum and urine)

were then developed. Whole deproteinated and extracted

samples of whole serum were analyzed to determine the best

sample form to use in terms of speed of analysis, sensitivity

and selectivity.









Instrumentation

The instrumentation used in this study was a Finnigan

MAT Triple Stage Quadrupole mass spectrometer (Figure 1)

with an INCOS data system (30). It consists of, in series,

a dual chemical ionization/electron impact (CI/EI) sample

ionization source, a quadrupole mass filter, an RF-only

quadrupole that can be pressurized with a collision gas,

a second mass filter, and an electron multiplier. The mass

spectrometer Was operated in the positive ion chemical

ionization mode using 99.9% pure methane (Matheson) as

reagent gas. Methane pressure in the ionizer was 0.30 torr

when using the Finnigan 4000 ionization source and 1.00 torr

when using the Finnigan 4500 ionization source. This reagent

gas pressure gave the best overall ionization efficiency for

large numbers of drugs; however, a higher or lower reagent

gas pressure may be used to give better ionization efficiency

when conditions are optimized for individual drugs.

All samples were placed in glass sample vials and inserted

into the ion source on the tip of a heated solids probe. The

probe was then heated ballistically from 30 C to 325 C in

approximately 100 seconds, and then cooled with Freon before

removing it from the vacuum. Up to fifteen samples were

analyzed per hour and the sample reproducibility (standard

deviation) was 15 to 20%.

Zero grade nitrogen (Matheson) was used as the collision

gas at a pressure of 1.8 millitorr in the center quadrupole.

An ion energy of 18 eV was used during all SRM analyses and

the collection of daughter spectra.























z- zo




Ps u
U I- w (


50
C c0
S| ZC









li L
ooH











0
0 gL
D I a rI-- < cc^ ^



cc

I I I







0 W I r-


g z 0R
S- 0r




0 m
<4J- WC





o 0 0






0fa





-- *H
--- W
Sd

3l

"1" ?
-* tr *
P-i









Reagents

All chemicals and drugs used in this study were reagent

grade and were used as received. A list of drugs and sup-

pliers is given in Table 1. The internal standards used

in this study, 2-amino-2-chlorobenzophenone and tribenzylamine,

were obtained from Aldrich and Chem Service, respectively.

Chloroform, ethyl ether and methanol were all Fisher reagents.

Acetone and sodium hydroxide were from Applied Science.

Equine blood serum and cannine blood serum and urine samples

were obtained from the University of Florida College of

Veterinary Medicine.

Sample Preparation

Deproteination

Serum samples were deproteinated by slowly adding an equal

volume of acetone to the sample while vortexing. The mixture

was then centrifuged and the aqueous layer decanted and

analyzed.

Serum Extraction

Acid-neutral components were extracted with 2 mL of ether

from 1 mL of serum acidified with 0.1 M HC1 to a pH of 3. The

ether was then evaporated to dryness and the residue taken up

in 1 mL of chloroform or methanol. The pH of the aqueous

layer was then adjusted to 10 with 0.1 M NaOH and the base

components extracted with 2 mL of chloroform. The volume was

then reduced to 1 mL.













Table 1
DRUGS, THERAPEUTIC CATEGORIES AND SUPPLIERS

Drug Therapeutic Category Supplier

Apomorphine Emetic, expectorant a

Aspirin Analgesic, antipyretic, b
antirheumatic anticoagulant

Caffeine Cardiac and respiratory b
stimulant, diuretic

Camphor Internally: stimulant and c
carmenative, Externally:
antipruretic, counterirritant
and antiseptic.

Chlorprontazine Antiematic, tranquilizer, d
sedative peripheral
vasodilation

Cocaine Local anesthetic, CNS stimulant e

Diethylcarbama- Anthelmintic, antimicrobilarial f


zine

Ephedrine


Fentanyl

Furosemide

Glycerol Guai-
acolate

Meperidine


Methadone

Methocarbamol

Methylphenidate
continued


Sympathomimetec, mydriatic,
CNS stimulant

Analgesic, tranquilizer

Diuretic, antihypertensive

Expectorant


Narcotic, sedative, analgesic,
anesthetic

Narcotic, analgesic

Skeletal muscle relaxant

Central stimulant









Table 1-continued

Morphine


Nikethamide

Para-aminobenzoic
acid

Pemoline

Phenacetin

Phenobarbital

Phenylbutazone

Procaine

Propylparaben


Reserpine

Theobromine


Theophylline


Narcotic analgesic, sedative <
preanesthetic, gastric sedative

Respiratory stimulant

Treatment eczema nasi in dogs


Central stimulant

Analgesic, antipyretic

Anticonvulsant, sedative

Analgesic, anti-inflammatory

Local anesthetic

Pharmaceutical aid (antifungal
preservative)

Hypotensive, tranquilizer

Diuretic, myocardial stimulant,
vasodilator

Diuretic, cardiac stimulant, i
vasodilator


a Merck Sharp & Dohme, West Point, PA
b Aldrich, Metuchen, NJ
c Chem Service, West Chester, PA
d R. J. Perchalski, Gainesville V. A. Hospital, FL
e Applied Science, State College, PA
f Sigma, St. Louis, MO
g Janesen Pharmaceutical, New Brunswick, NJ
h USV Pharmaceutical, Tuckahoe, NY
i Florida State Division of Parimutual Wagering, Miami, FL
j CIBA Pharmaceutical; Summit, NJ
k Dr. Jean DeGraeve, Universite De Liege, Belgium
1 Matheson-Coleman and Bell, East Rutherford, NJ
m Warner-Lambert, Ann Arbor, MI









Urine Extraction

Some urine samples were initially hydrolyzed by adding

an equal volume of concentrated HC1 to the sample and heat-

ing at 100 C for 20 minutes in the hood. These samples,

as well as unhydrolyzed urine samples, were extracted with

2 mL of ether from 1 mL of sample with a pH adjusted to 3

with 0.5 M NaOH or 0.1 M HC1. The ether was then evaporated

to dryness and the residue taken up in 0.5 mL of chloroform

or methanol. The pH of the urine was then adjusted to 10

with 0.1 M NaOH and the base components extracted with 2 mL

of chloroform. The volume was then reduced to 1 mL.

Animal Dosage and Sample Collection

Dosages of 200 mg of diethylcarbamazine, 500 mg of

procaine and 1 g of phenylbutazone were administered orally

to each of two greyhounds. Each animal was in good health

and weighed approximately 60 pounds. Samples of urine and

blood were taken from each dog prior to the administration

of the drugs. Blood samples were collected from each dog

after 1,2,3,4 and 6 hours. Urine from dog 1 was collected

at 2,3 and 5 hours and from dog 2 at 2 and 6 hours.















CHAPTER 3
RESULTS AND DISCUSSION

The rapid screening of complex mixtures such as blood

serum and urine for large numbers of trace substances by

tandem mass spectrometry is best accomplished using chemical

ionization in combination with selected reaction monitoring.

The use of chemical ionization, a more efficient and more

gentle ionization technique than electron impact, maximizes

the conversion of the sample to parent ions and minimizes

fragmentation. Although negative chemical ionization is

even more sensitive than positive chemical ionization for

many classes of compounds, positive chemical ionization is

most sensitive for most of the drugs that have been studied.

Furthermore, the stability of the negative ions also decreases

the efficiency of collisionally activated dissociation of the

parent ions in the center quadrupole, decreasing the sensi-

tivity of the screening technique.

The daughter ions which are monitored during SRM are

selected by first acquiring a complete daughter spectrum of

the pure drug's molecular ion. The most intense daughter ion

is monitored except in cases where two drugs have parent ions

of the same mass which fragment to give the same most intense

daughter ion. The most intense unique daughter ion from each

parent ion is then monitored. Figure 2 shows the daughter

































u
In

z




0I


N
w C
I
O=z



S ""


+
I:F


~D

In
4<--


o+
C---


d


4
+0)
ig________


z m
Z0
on O "
I 0 0



z 1


>- e
w C D
Io +
T ----: _


o
0
N--


.6


0
O
N
4--)






00

-0 -

C.
-,I
o










0






0 0

-0
c
0
























4-)
44
0
tll













u
0


























04


4-1
N C
u-P
















s


-VN 91S









spectra of the molecular ions of two drugs, procaine (237+)

and phenylbutazone (309+), from which the 100+ and 188+ ions

are selected as the daughter ions to be monitored with the

third quadrupole.

The fragmentation efficiency of the parent ion under the

conditions selected for the CAD process in this study varies

greatly depending on the structure of the drug. Ideally

the parent ion would fragment to give primarily one daughter

ion with a relative abundance similar to that of the remaining

parent ion in the daughter spectrum. This will give good

SRM screening sensitivity and still provide a daughter spectrum

with several characteristic daughter ions which can be used

as part of a standard library to confirm the existence of the

drug.

The rate of heating for the direct insertion probe was

chosen to give rapid desorption and maintain peak shape in

the "mass chromatograms" (plots of ion current vs scan number

or time). Using this heating rate, the most volatile compound,

nikethamide, desorbs off the probe almost immediately, while

reserpine desorbs only at 340 C.

Sample Preparation

The analysis of trace components in complex systems

usually requires a pre-analysis cleanup. This sample treatment

necessary to obtain useful analytical results directly affects

the analysis time, and thereby, the number of analyses that

can be performed.









Standard solutions of drugs in chloroform or methanol

were analyzed directly. The solvent was evaporated from

1 pL samples by warming them over the GC interface. A

mass chromatogram of the daughter spectrum data from the

237+ parent of procaine is shown in Figure 3(a). The peak

at 25 seconds represents 25 ng of procaine.

The analyses of serum samples were initially carried

out directly also. Refrigerated samples were allowed to

warm to about 70C. A 1 pL sample was placed in a glass

vial and evaporated almost to dryness over the GC interface.

The serum analysis yields a mass spectrum which shows the

presence of large numbers of mass peaks, one at almost every

mass. This is due to the thermal decomposition of the serum

which beginsat about 2900C. This decomposition causes a

rapid increase in the ion current as shown in the mass

chromatogram of a serum blank in Figure 3(b). In the direct

analysis of spiked serum, the mass chromatogram shows the

same increase in background due to the serum decomposition,

preceded by a peak indicating the presence of the added

component. Figure 3(c) shows the results of such an analysis

carried out on 25 ppm procaine in serum. The peak at 70

seconds corresponds to procaine, as confirmed by a comparison

of the daughter spectrum of procaine from 3(a) with the

background subtracted daughter spectrum from 3 (c). The two

daughter spectra are shown in Figure 4.












(a)


(b)


-_J I ../
SI I-
w (c)
------------ --------- ---- ~




(d)




25 50 75 100
TIME (SEC)
Figure 3: Mass chromatograms of (a) 25 ppm procaine in
chloroform, (b) a serum blank, (c) 25 ppm procaine
in fresh serum and (d) 25 ppm procaine in serum
which was frozen for seven days after procaine
addition.












25 PPM PROCAINE DAUGHTER SPECTRA

100


237


IN CHLOROFORM


120
1


164


150


200


Figure 4:


Comparison of the daughter spectra of procaine
from (a) Figure 3(a), and (b) Figure 3(c).


100.





50-


50






M/Z


IN SERUM


250


I I I II I









The 45 second shift in the procaine desorption time

from the chloroform solution to the spiked serum analysis

is due to protein binding. At low concentrations, essentially

all of the procaine is bound up in the serum protein and it

is not released until the protein begin to thermally decompose.

At higher concentrations (100 ppm), procaine is present in

both the free (unbound) and bound forms as shown in Figure

3(d). This indicates that studies to determine the ratio of

protein bound to free drug could be performed by MS/MS.

Experiments were carried out to determine the method

which could most easily and efficiently be used on serum

samples to free bound drugs prior to analysis and to decrease

the background due to thermal decomposition of the serum. This

would also serve to allow more analyses to be performed before

needing to clean the ionizer or change the ion volume.

Only 50 to 60 serum samples could be run before a significant

decrease in sensitivity (due to deposits from the serum in the

ion volume) was noted.

Direct analysis of urine samples was also carried out.

Interactions between the drugs and the components of the.urine

were much weaker than those in blood serum allowing procaine

to desorb after 50 seconds. There was little thermal de-

composition of the urine evident in the ion current trace over

the temperature range used, and, therefore, little increase

in the background. The direct analysis of urine, however,

allowed only 35 to 40 samples to be analyzed before a signi-

ficant loss of sensitivity was apparent.









The simplest apparent solution to the problem of protein

binding was to denature .the protein and free the drugs which

were bound to it. Several compounds, including trichloro-

acetic acid solution, methanol and acetone were evaluated as

denaturing reagents for the serum proteins.

Trichloroacetic acid as well as methanol and several other

reagents gave small globular precipitates which seemed to

trap much of the bound drugs inside. Analysis of the supernat-

ant solution showed recoveries of 30 to 40%. A much looser,

more fluffy precipitate was formed with acetone, and analysis

showed recoveries of 45 to 55%. Although denaturing the

proteins with acetone gave better recovery and greatly

decreased the background at high temperatures, the recovery

of the bound drugs was still low. Concentration of the super-

natant. solution was also a comparatively slow process, so

that the analysis for drugs using deproteination is limited

to systems where analyses of high concentrations of drugs

in serum is desired.

Modified procedures for acid-neutral and basic extractions

were carried out on spiked serum samples. Recoveries of

70 to 80% were observed, a significant improvement over those

obtained from denaturing serum. The use of ether and chloro-

form to extract the acid-neutrals and bases, respectively,

also allows for a rapid concentration of the drug containing

organic phase to volumes as low as 50 VL. The extraction of

larger serum volumes can improve the drug detection limits









in serum even more. Figure 5 shows a mass chromatogram

from the SRM analysis of the base extract of 2 mL of serum

containing 4 ppb procaine. The extract was concentrated

to 0.1 mL and a 1 pL sample analyzed.

The current procedures for the extraction of drugs and

metabolites from urine call for acid or enzyme hydrolysis

before the extraction (3,4). Urine samples in this study

were hydrolyzed by adding an equal volume of concentrated

HC1 and heating. These samples were then extracted using the

same procedure that was used for the extraction of serum

samples. Drug recoveries were about 80%. Analyses performed

on these samples to determine the metabolites of several drugs

indicated that some of the metabolites had also been hydrolyzed..

Urine samples were then extracted as before, but without

prior acid hydrolysis. Drug recoveries were comparable to

those obtained from the hydrolyzed urine for the three drugs

under study, diethylcarbamazine, procaine, and phenylbutazone.

Analysis of this extract showed the presence of several meta-

bolites which were not present in the hydrolyzed urine extract.

The metabolites in question were all oxidation products of

the parent drugs. These metabolites appear to have undergone

dehydration during the urine hydrolysis. This seems to

indicate that a less concentrated acid should be used for the

hydrolysis of the urine samples. Subsequent urine analyses

were carried out without hydrolysis since extraction recoveries

for the drugs were equivalent and extraction without hydrolysis

was more rapid.















0
0
0

&,


0o
*'




o




0 )-


4
o


















x



4-1
44



0
Sa
-d








C) r-
a)
O


--I




0
c




Cso



4


0


u) (0























E*d









Screening

Screening,as currently employed, is used to detect the

presence of any of a large number of trace constituents in

a complex mixture such as blood or urine. General methods

which can identify drug classes or functional groups are

used as well as tests developed to identify specific drugs.

The procedures are designed to be as fast and as inexpensive

as possible while maintaining a selectivity and sensitivity

that can be used to control the use of illegal drugs in

racing animals.

Tandem mass spectrometry is ideally suited for the analysis

of selected compounds in a complex mixture. The first mass

filter is used to filter out all masses except that of the

compound of interest. All ions having that selected mass

are then fragmented in the collision chamber and the daughter

spectra collected. When SRM is used, all available time is

spent looking at the signal, or signals, of interest. No

time is spent in the collection of background data while

scanning from peak to peak. Scan times over each single mass

unit being monitored are 0.1 second or less. This allows

sequential scanning of large numbers of parent ion/daughter

ion pairs during an analysis. The number of parent ion/

daughter ion pairs is limited only by the sampling interval

required for each compound. The rate at which the direct

insertion probe is heated and the corresponding peak widths

for each compound as it is desorbed has a direct bearing on

the sampling interval required. Typical peak widths are








about 10 seconds. If four sampling intervals are desired

for each compound being monitored, 25 different parent ion/

daughter ion pairs can be sampled at 2.5 second intervals.

This rapid, selective type of analysis was used to screen

complex mixtures for trace amounts of drugs and for

metabolites.

To demonstrate the potential of SRM as a rapid, selective

screening technique, a mixture of two isomers theobrominee

and theophylline) and a third compound (propylparaben) with

the same nominal molecular weight of 180 was examined (Figure

6). The daughter ions 124 138+ and 139+ were monitored

for the compounds theophylline, theobromine and propylparaben,

respectively, while 181+ parent ions were passed through the

first quadrupole. The resulting mass chromatograms are shown

in Figure 7. Each compound shows a distinct peak which is

slightly offset from the others, indicating some separation

arising from the thermal distillation of the mixture off the

solids probe. The bottom trace shows the ion current produced

by monitoring the 181+ parent ion, the equivalent of single

ion monitoring for this mixture.

The ability of tandem mass spectrometry to identify trace

amounts of selected compounds in a complex mixture was

demonstrated by adding 30 ppm theobromine to equine blood

serum. A conventional chemical ionization mass spectrum of

this mixture is shown in Figure 8. The thermal decomposition

of the blood serum at about 290 C produced peaks at almost












CH3 0
N N/H

N N 0N

CH3
THEOBROMINE


H-0


N CH3
KI N 03

CH3
THEOPHYLLINE


-O-CH
C-0-C3H7


PROPYLPARABEN

Figure 6: Structures of (a) theobromine, (b) theophylline
and (c) propylparaben.




















Co -


4-)


CC 4C

z
0I
S-4,
N
+









" ap o0

0 .4-
00
-i--
L) E 0

CC M



0o
ao o o tn










n

4 4
*H C


*ir











0)04
*4O 0



o
at- ,








I,-I
Q) U 4-)









4'


liJ




T
0
I
-z
I-


0
-z
-=E
SO
To0
m
_O
.--





I


O
aD


~s~C~









every mass, including mass 181, which effectively masked

the presence of the trace component. The same mixture was

then analyzed using the same SRM conditions as were used in

the analysis of the three component mixture discussed above.

The resulting mass chromatograms (Figure 9) show that the

theobromine was detected. The mass 181 trace shows two peaks,

one at 70 seconds and one at 100 seconds. The peak at 70

seconds shows the presence of theobromine, a parent ion of

mass 181 which fragments to give a 138+ daughter ion. The

latter peak, which occurs at all monitored masses, is an

increase in background due to the fragmentation of the 181+

ion from the thermal decomposition of the serum. Protein

binding, as discussed earlier, shifts the theobromine peak

from 25 seconds in chloroform (Figure 7) to 70 seconds in

blood serum (Figure 9). There is also an increase in the

detection limits due to the increase in the background from

the thermal decomposition of the serum which overlaps the

desorption of the theobromine.

The use of SRM for screening complex mixtures can, in

some cases, reduce the background from the thermal decomposi-

tion of serum which was present in the daughter spectrum.

Figure 10 shows the mass chromatogram acquired while using

SRM to screen a serum sample containing 25 ppm procaine.

The only peak present is from the desorption of procaine at

70 seconds. The background increase which was seen in the

daughter spectrum (Figure 3) is no longer present. The





































































7-VN9IS


*d



*a)



02
0
C



e-I

4-4

r-I

rw
0



-4
0

4-1


04
wo
O




e.




0
Cn
*. I *


o)


?i







(-S


w
z
OD


Ts
mI
.. 0
OD
--w


0,

I`

0o










T













4J
0

0




-1









0

,-1
















0


o
Lr










a








r o
(U
0I



0)




S4 (0





0 0
1-1
&w











0~


1 V N IS









fragmentation of the serum component with a parent ion of mass

237 did not produce a fragment with mass 100, allowing the mass

100 fragment of procaine to be monitored without interference.

A mixture of four compounds (theophylline, propylparaben,

phenylbutazone and procaine) in serum was extracted and analyzed

using the SRM screening technique. The results of screening

the base extract are shown in Figure 11. Procaine, the only

base, gives a distinct peak in its mass chromatogram indicating

the compound's presence. The mass chromatograms of the acids

and neutrals, theophylline, propylparaben and phenylbutazone,

show only low intensity background indicating their absence

from the base extract. Each point shown in the mass chromato-

grams represents the ion current collected in a 0.1 second

scan over the indicated mass units, so that each of the four

compounds is being scanned for 25% of the available analysis

time.

So that the screening procedure which had been developed

could be applied to some real samples, serum and urine were

obtained from two greyhounds. The dogs had been dosed with a

daily therapeutic dose of diethylcarbamazine for four days,

and, on the fourth day, with a single therapeutic dose of

procaine and phenylbutazone. Whole serum samples were screened

first to test the feasibility of analyzing samples without

some preliminary cleanup, followed by the screening of the

acid-neutral and base extracts of the serum.

The two bases, diethylcarbamazine and procaine, were both

present at very low levels, procaine at about 70 ppb and





















w
z



I


-w
a.

H-


*

ID
C\J

IVN9IS


0
0
- z



'ocr
Na
tx I1


z
w




-mo
crr

a,



, -
CL
OD0
1s .


w


OD
T z





0)
m *


ow
I

*


. o


z


cov


oocc


Cc,
5)


x0
4-,
j-l








O
x 0


WO

*H
co





(U
4 -1
+- 4-)




























00
c0
0 U
tn














r4
41

-I
Ur(






ca


0 0




d H
(0
En

0C


N0


0
*








diethylcarbamazine at about 400 ppb. The low concentrations,

in combination with the protein-binding and increasing back-

ground due to thermal decomposition of the protein, made

the identification of the two compounds difficult in whole

serum. Confirmation by library comparison of daughter spectra

was almost impossible. The mass chromatograms are shown in

Figure 12. Both diethylcarbamazine and procaine were desorbed

almost totally under the leading edge of the background from

the thermal decomposition of the serum. The level of the

background in the daughter spectra could not be sufficiently

reduced by background subtraction to obtain a daughter spectrum

that could be matched with the library spectra. Most of the

phenylbutazone was desorbed prior to the thermal decomposition

of the serum and the background subtracted daughter spectrum

could easily be used to confirm its presence in the serum

(Figure 13). Based on these analyses, it appears that the

direct analysis for most bases, which normally occur at very

low concentrations in serum, will not be possible under these

conditions. The direct analysis of acids, however, appears

to work well since acid concentrations in the serum are

generally 100 to 1000 times greater than base concentrations.

One mL samples of serum were extracted and screened using

the same procedures as were used in the analysis of the whole

serum. The mass chromatograms from the screening of the serum














U)


aP
0
()1






,L,
pN






/ \

3'-


oc

0 O


z ( I ) N-
--, -^










.-
z a





g ^s



o 0
ei


,'-I
0


04




-NDIS C

ori




(4


0
0

T

h-
Od

f
N








D















































































































e


* rJ





I-








'.
NOD)CQ

0;5~I
.O OI=-MOIO








extracts show that both diethylcarbamazine (Figure 14 (a))

and procaine (Figure 14 (b)) are desorbed much earlier and

are more easily identified in the extracts. The phenylbutazone

(Figure 14 (c)) shows a very intense signal indicating a high

concentration in the serum.

Daughter spectra were collected for each compound and

compared to the library of CAD standards. The library matches

for diethylcarbamazine, procaine and phenylbutazone (Figures

15-17) show excellent agreement for diethylcarbamazine and

phenylbutazone and leave no doubt that they are in the serum.

The low concentration of procaine, however, makes it difficult

to successfully identify. Extraction of a larger volume of

serum or concentration of the extract to a smaller volume could

be used to increase the procaine signal intensity and make

identification more definite.

From this comparison, it is evident that the use of an

acid-neutral and basic extraction step in the screening

procedure facilitates screening for the drugs and makes

confirmation of indicated positives more likely. The additional

sample preparation time is short, and, it allows an increased

number of analyses to be performed before a loss of sensitivity

due to source contamination.

Canine urine samples were analyzed using the same format

as was used for the serum analysis. The urine itself was

screened first, followed by acid-neutral and base extracts.

The screening of the whole urine gave surprisingly good results.














()

4
4-
5-4
0

4-
0

0. -
u
4J

4-)



5-

4n

a





0







IC



4'



4-)

u c


0






U
C

0d


"IV NIS



















0




N




E e




ro








H a
0




U







a
--0
X4


Cd


4J





AiI-i



.,.


4- )
e






-CO1

U-)





5 (=0
+>namIn N




























































































CM







(a---*-04

M0< ND C<1

2-J
U; Sfas-


O0

00
C. Y
2 a
0)


rT4






48

















o
O
0
t4c

a4


N

4 +J
















a-
-i














04
U E










-I
=- ~



















0
U
ad
















0











.a
Q4
r-
- ne-m








( rc














Ere








The mass chromatograms shown in Figure 18 show very strong

signals for all three compounds, particularly the diethyl-

carbamazine and procaine. The phenylbutazone, even though

given in the highest dosage and appearing in the highest

concentration in the serum, is less than 3% of the signal

of either base. The positions of the three peaks are shifted

only slightly from the standards and daughter spectra of the

three compounds easily match the library CAD spectra. The

low base concentrations in serum and high concentrations in

urine indicate that bases are rapidly removed from the blood

and excreted from the body in the urine. The acids, however,

remain in the body longer, yielding higher serum concentrations.

The same relative signals were .obtained from

screening the urine extracts (Figure 19). The signal inten-

sities are somewhat less due to extraction recoveries of only

70 to 80%. This suggests that the direct analysis of urine

samples will provide better sensitivity than the analysis of

the urine extracts. The main limitation to the use of urine

in the screening was the comparatively rapid rate at which

sensitivity was lost when repeated analyses were performed.

This problem has become less important, however, with the

use of the Finnigan 4500 ion source which has an easily

replaceable ion volume.

Confirmation of the positives indicated in the screening

of the urine samples was carried out by collecting full

daughter spectra of the three components. The library matches
















t.









O
Q)



0
4N
Or
0 (U

















-,-'
0





Q
0




4-)
UC r-



N
cH











o c
o*





00
N


U H
O1 0
(U-c-
E; .







(U (





















i ON



O) .(
4J a

CdrH




O(
C-**
-4 >


l OC

S,--)
U C

*H



0 0

0







CN'
4-I





.4
U--


4 N



O




oE
Eno




r i
(IO

O ..4









:d



0*.



41


'H









for diethylcarbamazine, procaine and phenylbutazone are

shown in Figures 20-22. The matches with the reference

spectra are very good. Daughter spectra collected from the

urine extracts gave comparable library matches.

The total number of drugs that might be illegally

administered to racing animals is very large. In some states

several thousand compounds are classified as illegal if found

in the system of racing animals. Screening for each of these

drugs individually would be a very difficult task. The total

number of analyses that must be performed can be greatly

reduced by combining compounds into groups, or classes, which

have structural similarities. This is possible by screening

for a given neutral fragment loss which is characteristic of

several compounds, or looking for all compounds which give a

selected fragment. For example, nikethamide, diethylcarbamazine

and procaine each contain a terminal -N(C2H5)2. When any of

these three compounds is fragmented, it looses diethylamine

(mass 73) as a neutral molecule. Figure 23 shows the results

of screening a mixture of seven compounds, nikethamide,

diethylcarbamazine and procaine as well as ephedrine, methyl-

phenidate, meperidine and methadone for a neutral loss of 73.

The resulting mass spectrum shows peaks at 106, 127 and 164

representing the (MH-73)+ ion from nikethamide, diethyl-

carbamazine and procaine, respectively.

The canine urine extracts analyzed earlier by screening

for individual compounds, were screened for a neutral loss



































































Ir










r










a,
=x


(U




On
.4 0

I



N



-Q





c.

-H


o
0







O
0






*1
oo
0












4J
ru












5I



>-I




C)
41
0



*4


L


J
1
-C
a,









































































*C
a











N
^c o--r
COe IBS _


L

I.
Ct


Cr,
<-. .1

fcD
qJ


a,

4 -








ol-











44
0
Cd
U


4-I














,-
0






s-I

4-4





a,


s
-Q


cU

















a)
5.

-'-
















































































I.!
-------1~-g!:--T-c----
0 *dS8-"a
to M jocn o

"I ,S'si^
to u :::ae-u.


CM


tp
H)

4
Cnl


*-



0





N
4-)

Cr

0

E,
Q)
c












o o










'-1
0


















14
c,
U
c)


4-,



C*
. )
+d




(U
4-
kS t
3



















































































-1VNEIS


ID------









NC


o)





0 W
4-) >)
MyO
H
m


.4 r-)
o (0
,--4 M 5.
+ 40






m
0 C H






Eoo
,-4 r
4J 4- 0


- m
Qo :l


. o
-n O4
E4 C U
O- o







0 L0



ow




04
Cd
*0 )
5-1













14
Sl i
rm *




}^ (l


C,)
- C
V)



z
0
Qm

rw



-0












O
0









N
2S


rN-------------- -








of 73 (Figure 24). Mass peaks at 127 and 164, the two

most intense peaks in the spectrum, represent diethylcarbama-

zine and procaine. A third peak at mass 113 was initially

thought to be a natural constituent of the urine; after a

few calculations, it was realized that it was actually an

N-dealkylation metabolite of diethylcarbamazine with an

MH of 186. These positives from the neutral loss screening

couId b-e confirmed just as positives are identified using single

component screening. The daughter spectrum of each MH+ ion would

be collected and compared to the library of standards.

In addition to screening for neutral losses, screening

for all members of any class of drugs that dissociate to

produce a common fragment can be carried out by using the

parent spectrum operating mode. For example, three xanthines

(caffeine, theobromine and theophylline) each fragment to

give a 163 ion. By scanning quadrupole one and looking for

the 163 ion with quadrupole three, the parent ions (MH )

181 from theobromine and theophylline, and 194+ from

caffeine appear as peaks in the mass spectrum. The use of

parent and neutral loss screening can greatly increase the

number of drugs that can be detected during the analysis

of a single serum or urine sample.

The confirmation of the existence of a selected compound

is necessary following any indication of its presence during

the screening procedure. Although two mass separations

are used in the screening procedure to gain a high degree


















*0
'N
0
4"-










4a>

0
CdC
fo









r. (a

4-J
--- I





+ o4
0

4-




0 (
r (









0 =
00











4- !
co l T




























04
L-4






LO
0U 0
0o14
'-2 on









z *












| -- i -- i -- -- i -- 1 -- 1 -- -- i -- i -'^


G. W
G> _S
0 Ul 5









of selectivity, there is still the possibility of more

than one compound of the same mass producing fragments of

the same mass. The collection of a full daughter spectrum

of the screening positive allows it to be compared to

standard daughter spectra and identified. Anticipated

problems arise when drugs which have inactive d or 1 isomers,

such as levorphan (active) and dextrorphan (inactive), are

suspected after screening has indicated a positive. This

may require the useof achiral CIreagent gas which is selective

for either the d or 1 form of the compound (31), or deriviti-

zation of one form of the compound. In general, however,

the matching of daughter spectra obtained from complex mixtures

following a screening positive will be sufficient proof of

the drugs'presence.

The library comparisons are carried out by the INCOS data

system. Figure 25 shows a printout of the results of a

search of library PX for a component from canine urine. The

five best matches from the library of 42 compounds, ranked

by fit, are listed along with their molecular weight and

base peak. The sample daughter spectrum can be visually

compared with any of these daughter spectra. Figure 26 gives

a display of the daughter spectrum of the sample and the best

match. The differences between the two spectra are highlighted

below.

Currently there is no library of daughter spectra available

for general use. One of the main problems is promoting a














MID LIBRARY SEARCH DATA:
84/22/82 14:59:08 + 8:32 CALI:
SAMPLE: DOG 1 MIX 5 HR
CONDS.: SOLIDS PROBE
* 26 TO a 48 SUMMED # 1 TO 0


U1X3DS%:B082 # 37 BASE M/E: 189
CAL0422Q3 0 1 RIC: 281599.


21 14 X1.08


42 SPECTRA IN LIBRARYPX SEARCHED FOR MAXIMUM FIT
27 MATCHED AT LEAST 1 OF THE 16 LARGEST PEAKS IN


THE UNKNOWN


NAME
PROCAINE
PARA-AMINOBENZOIC ACID
2-METHYL-2-PHENYLSUCCINIMIDE
PHENACETIN(ACETOPHENETIDE)
CAFFEINE


RANK FORMULA
1 C13.H28.02.N2
2 C7.H7.02.N
3 C11.H11.02.N
4 C18.H13.02.N
5 CS.H10.02.N4


M. UT
236
137
189
179
194


B.PK
100
138
120
138
195


PURITY
787
58
51
17
30


Figure 25: Computer listing of the results of a
search showing the five best matches
by fit.


RANK
1
2
3
4
5


FIT
947
460
449
447
489


RFIT
791
78
98
19
34


library
ranked








































































m





w w
+ Lnl
Cu-












L .0
(COD ** -r





c tn l-l



5--
v>:


I



f. -

C
C.------
Qr C4


H

-I.



"4I
44







-l
44

C
4-)











0
( 1





rd
s-1





















4O
sa-









4m

0
U




















0CQ
Cd








1 -










4-)
(0 *
*J4 6t
ow


Cd-2
1) +
.QWU
C <


_ __ ___









consensus on the standard conditions to be used in such a

library. The library constructed during this study used

the conditions listed in the experimental section.

Methanol solutions of pure compounds at approximately

50 ppm were used as standards. A conventional chemical

ionization mass spectrum was first collected to check purity,

determine volatility and determine the transmission of the

parent ion. The daughter spectrum of the compound was then

collected and inserted into a library. The daughter spectra

which are currently contained in this library are listed in

Appendix A.

The computer that controls the triple quadrupole mass

spectrometer also provides a convenient means by which the

screening can be automated. With the currently available

software, a procedure can be written which collects the

desired data (SRM, neutral loss or parent screening) and then

splits the file into separate files for the identification

of individual compounds or groups. If a peak is present in

a file, a daughter spectrum can be automatically collected

and compared to the library spectra. The resulting matches

can be printed for operator evaluation.

The software makes it possible for the computer to change

operating modes, collision energy and multiplier gain for

each scan as well as to turn the filament and electron multi-

plier on or off. This versatility makes the instrument

ideally suited to analyze and evaluate data quickly,. waiting









only for the operator to prepare the next sample for analysis.

An experiment written in order to screen a complex mixture

for nikethamide, diethylcarbamazine, procaine, caffeine, theo-

bromine, theophylline, methadone, apomorphine, meperidine,

methylphenidate and ephedrine is depicted in Figure 27. Five

of the drugs are sought individually using SRM screening

while three (caffeine, theobromine and theophylline) are

screened by a parent spectrum of 163+. A neutral loss of 73

is used to indicate the presence of nikethamide, diethyl-

carbamazine and procaine. Scans, as described by each :

successive entry in the experiment, are collected in sequence,

top to bottom, repeating until the requested number of scans,

or time interval, has been completed. The resulting data file

is split into seven separate files, each containing informa-

tion on one drug or drug class. These data files are then

evaluated to determine the possible presence of any of the

drugs.

The use of the system computer to carry out the screening

and data analysis automatically is a logical step in the develop-

ment of a rapid analytical procedure. It will minimize operator

interaction and leave the operator to load samples and evaluate

the results. A procedure to screen for selected drugs, split

the file and evaluate the results and one to collect the

daughter spectra and compare them with the library spectra

were written. These procedures- can be tied together to enable

the system to control the analysis.




















EXPERIMENT EDITOR


CALIBRATION TABLES:
SYSTEM DESCRIPTORS:


INPUT FILE:HO


QUAD ONE
CAL8719Q1
01


QUAD THREE
CAL071903
Q3


7 MASSES BETWEEN 73.88 AND 318.08
7 MID DESCRIPTORS: NL.P1.D1.D2.D3.D4,D5
7 CONFIG DESCRIPTORS: NL73.P163,DDMX.


ENTRY MASS MID TIME
NO. SET DESC. (SEC)


73.68
163.08
166.00
234.00
249.88
268.88
318.00


(03)
(01)
(Q3)
(03)
(03)
(03)
(03)


8.19
0.17
0.11
8.11
0.11
0.11
0.14


MASS CONFIG.
WINDOWS DESC.


50-
50-
147-
83-
173-
236-
264-


501
451
149
85
175
238
266


NL73:NEUTRAL LOSS/+/GAIN:7/CE:-17.9
P163:PAREITS/+/GAIN:7/CE:-17.9
DDMX: DAUGHTERS/+/GAIN:7/CE: -17.9
DDMX:DAUGHTERS/+/GAIN:7/CE:-17.9
DDMX:DAUGHTERS/+/GAIN:7/CE:-17.9
DDMX:DAUGHTERS/+/GAIN:7/CE:-17.9
DDMX:DAUGHTERS/+/GAIN:7/CE:-17.9


Figure 27: Computer listing of an experiment written to
screen for a neutral loss of 73, parents of
1.63+ and five individual compounds.









Quantitation

In general, quantitation is infrequently employed in

the analysis of drugs in racing animals, as the detection

itself is sufficient to bring about any legal consequences.

Quantitation of trace amounts of drugs in serum and urine

is widely used, however, in metabolic and pharmacokinetic

studies. The most common method of quantitation for these

studies involves the use of isotopically labeled internal

standards (32). For the analysis of small numbers of compounds,

the purchase or synthesis of labeled internal standards may

be practical; for screening for large numbers of drugs it

can be very expensive, or very time consuming. An alterna-

tive to the use of labeled internal standards, the use of

sample-independent internal standards, has been proposed by

Heresch et al. (32).

A sample-independent internal standard can be used to

quantitate a single compound, but it is ideally suited for

the analysis of multicomponent systems. Once the ratio of

sample and internal standard signals vs concentration has

been determined for drugs and metabolites of interest, the

internal standard can be added to all samples as they are

analyzed and drugs which are present can be quantitated. The

addition of the internal standard parent ion/daughter ion

pair to the experiment being carried out allows the quantita-

tion of drugs in the sample.









A large number of compounds were evaluated as internal

standards. Since acid-neutral and basic extractions were to

be used in the screening and analysis, two internal standards

were selected, one which extracts with the acid-neutrals and

one which extracts with the bases. They were 2-amino-5-chloro-

benzophenone and tribenzylamine, respectively. The parent ion

of each of the internal standards fragments to give one major

daughter ion: 232 + 154+ for 2-amino-5-chlorobenzophenone

and 288 +- 91 for tribenzylamine. A plot of the ratio of

the sample and internal standard signals vs concentration for

each of three drugs (theophylline, theobromine and phenyl-

butazone) shown in Figure 28, is linear from the low ppm range

to concentrations approaching 200 ppm using a 25 ppm concentra-

tion of internal standard. The compounds were mixed into

equine serum, extracted and the extract volume brought to the

original volume of the serum. The recovery of the compounds

from serum was approximately 80%.

A mixture of five drugs (theophylline, theobromine, phenyl-

butazone, propylparaben and procaine) and 2-amino-5-chloro-

benzophenone in serum was extracted and analyzed using the

SRM screening procedure. The results of screening the acid-

neutral extract are shown in Figure 29. Using these signals,

the concentrations of the extracts were determined to be

55,58,55 and 100 ppm for theobromine, phenylbutazone, propyl-

paraben and theophylline, respectively. These values vary

less than 10% from the actual concentrations (52,55,50 and

102 ppm, respectively) added to the serum.





67





,0 ,.
Cd







O 0


0 -4o

\ r-- (U










co
0g



0z 4
0 Q 0








4
O tP

S -4 r-C
a ,rl .
SZ
) a) -
\ Cu
\\ \-4 -'l









\ \ r-a,













3-dYS4V Crl
'\P~ *n 0 t>
\\ ^ *<-1
\^ -l D





68











x




0 0




u
Sz N





0 w Iz
*. z- *
O
S O .* X




_
w C N



u c 0 n 44
0 0



I *

I. 0 4
S* I
-* I* .0 2 ()


S -


." .I
.- o i. Q)





o. : I I m a
. *. ? :j : 0.0 -
. : 4-)
* ." C U

*. : -
<. u ( o3











: -




":: 0 N


E OS



0~>



S* N' 0 IS




C*H



rj
0m
'; : -1 C
*~~~u .d '*
*;~ :: : 5 '*
;* ^ (
.
; i
'. 03 4-
0 2; o v r -- r. -- s
E ^ ~ S
-'~~ 0> 0





~tlNBI-1-1









The detection limits for several drugs obtained with

the SRM screening procedure are compared in Table 2. The

wide range of detection limits in chloroform is primarily

due to two factors: the ease with which a compound is

ionized using positive CI, and the degree of fragmentation

that occurs in the center quadrupole under the conditions

selected. Procaine, the compound with the lowest detection

limits, is easily ionized and fragments so that a daughter

ion is the base peak in the daughter spectrum. Morphine,

on the other hand, does not fragment easily, the most intense

daughter ion being less than 5% of the parent ion. This

increases its detection limits since most of the morphine is

still molecular ion and is not being monitored.

The detection limits for most compounds are 2 to 40 times

higher in whole serum than in chloroform. This is due to

protein binding and the background produced by the thermal

decomposition of the serum. The addition of a simple acid-

base extraction scheme to the screening procedure reduces

the detection limits into the range of those obtained for

pure drugs in chloroform. The use of extractions as part

of the analytical procedure is particularly important when

screening base extracts due to the low levels at which the

bases are present in serum. Acidic drugs can usually be

screened in serum itself.













Table 2
DETECTION LIMITS FOR SIX DRUGS IN
BY SELECTED REACTION MONITC

CHC13


Procaine

Phenylbutazone

Para-aminobenzoic Acid

Theophylline

Theobromine

Morphine


30 ppb

800 ppb

1 ppm

2 ppm

10 ppm

70 ppm


1 UL SAMPLES
)RING

Blood Serum

Whole Extracted

4 ppm 100 ppb

2 ppm 1 ppm

30 ppm 3 ppm

16 ppm 8 ppm

20 ppm 14 ppm

90 ppm 75 ppm









Metabolite and Pharmacokinetic Studies

The identification of metabolites in serum and urine is

a challenging problem for medical and pharmacological

researchers. The concentration of metabolites in serum is

usually quite low due to the low concentration of the drugs

themselves and the rapid removal of the metabolites from

the blood by the body. This limits the ability to detect

the metabolites in serum without extensive sample preparation

and concentration. Metabolite concentrations in urine, however,

are significantly higher due to the concentrating effect of

the constant removal of the metabolites from the blood. This

simplifies the procedures needed to identify any metabolites.

A procedure which should significantly reduce the time

and effort currently required to find and identify metabolites

(29) was used to identify the metabolites of phenylbutazone,

procaine and diethylcarbamazine. Serum and urine samples from

greyhounds that had been dosed with the drugs were analyzed.

The major fragments of each of the drugs were determined by

collecting CI mass spectra of the pure compounds. A daughter

spectrum of the MH+ ion and each major fragment was then

acquired. From these daughter spectra, the masses of the

structural subunits of the drug were determined. Parent spectra

of these ions were then used to identify any ions in the serum

or urine extracts which fragment to give one or more of the

structural subunits of the drug. A daughter spectrum of

each of these probable metabolites was then collected and

evaluated to determine if it was a metabolite.









Serum

The structural subunit masses of diethylcarbamazine,

procaine and phenylbutazone are shown in Table 3. Parent

spectra of the structural subunits of diethylcarbamazine

were collected from a sample of serum extract. The resulting

masses of possible metabolites are shown in Table 4. The

(M+29)+ and (M+41)+ ions of mass 228 and 240 are present in

the parent spectra of the 200 198+ and 100+ ions. Parents

of mass 212,213,215 and 255 proved to be unrelated to the

diethylcarbamazine based on an examination of their daughter

spectra. The 127 ion is primarily a fragment of a hydro-

carbon chain as indicated by the sequence of parent ions

obtained. The presence of any metabolite fragments is effect-

ively masked by the hydrocarbon fragments present. The 100+

ion shows two additional ions of mass 186 and 200. The

daughter spectra of these two ions showed the mass 200 parent

to be diethylcarbamazine and the mass 186 parent to be an

N-dealkylation metabolite of diethylcarbamazine. The daughter

spectrum shows a mass 113 peak has replaced the mass 127 peak

from the diethylcarbamazine, indicating a loss of the methyl

group from the piperazine ring.

Parent spectra of the structural subunits of procaine

from Table 3 were collected from a sample of serum extract.

The resulting masses of possible metabolites are shown in

Table 5. The (M+29)+ and(M+41)+ ions of mass 265 and 277

are present as parentsof several ions. The mass 100 parents






/j
o

o N



00

r- -- 0
0 Oo '.0 (NC

N ( o

< C
0 0 C O




0 H C
i1 CO r-


S- 0 CM 00

S0 0 C0
04 ( N 0m H H
Z M

0 O -
cH CM ko H (N q
H H0 'OD CM 1-N H


H -(
H J O M H CO H A LA


U O H I'0 C0 k.0

H "I HA CN r-H CN
P4





H


n + OD LA
w v ( '.0








H 0 O 0 0 a
0C C> C0 0 0 C0


0 0> M rH 0 0
C) CQ Y ooe an C



(N CN N M n




P4


0 4-) 4- 4-0

P 4C CM CH H
)u u u)









r4 (o (


E4 P4



U4 P4
tr n, 1













Table 4
PARENT IONS OF DIETHYLCARBAMAZINE FRAGMENTS

Fragment
Mass Parents (abundance, %)

.ract 200(100) 215(16),228(26),240(10),255(2

198(100) 213(28,228(15),240(25)

127(48) 183 (35),197(38),211(40),225(5

100(55) 186(60),200(100),212(60),228(


Urine Extract


200(100)

198(100)

127(13)


100(8)

72(11)

113(20)


5)



0),239(72)

40), 24 0(70)


228 (22),240(8)

228 (20),240(28)

142(6),170(4),184(21),198(47),200(100),

228(20),240(10)

186(13),198(25),200(100),228(12)

184(23),186(48),198 (28),200(100),228(22)

226(28),214(38),212(11),200(12),186(100),

184(67),170(34),141(9)


Sample

Serum Ext
















Sample

Serum Extract


Urine Extract


Table 5
PARENT IONS OF PROCAINE FRAGMENTS

Fragment
Mass Parents (abundance, %)

237(100) 225(58),265(22),277(8)

164(100) 148(95),265(8)

120(98) 148(95),190(100),218(98)

100(12) 186(48),200(100),212(75),22


8(10)


237 (100) 255(48),265(42),277 (12)

164(19) 209(6),221(22),235(95),237(100),265(15)

120(100) 138(14),178(20),196(16),224(11),209(18)

221(13),235(25)

100(5) 186(13),198(25),200(100),228(12),237(18)









are identical to those obtained for diethylcarbamazine, and

show no relation to procaine. Daughter spectra of the

remaining parent ions also indicated none were related to

procaine, and were probably naturally occurring species in

the serum.

Parent spectra of the structural subunits of phenylbutazone

from Table 3 were collected from a sample of serum extract.

The resulting masses of possible metabolites are shown in

Table 6. Again, the (M+29)+ and (M+41)+ ions as well as the

MH ion (masses 337,349 and 309 respectively) are present

as parents for several ions. Daughter spectra of the mass

206 and 293 ions indicate they are not metabolites of phenyl-

butazone. The 325 parent ion indicates the presence of a

metabolite, an oxidation product, and the 353 parent ion is

the (M+29) ion of the metabolite.

In each case, the drug as well as its (M+29)+ and (M+41)+

ions were easily identified but information relating to any

metabolites was limited. Only a major metabolite of diethyl-

carbamazine was identified while none of procaine's metabolites

were detected. This is primarily due to the low concentration

of the basic drugs in serum. The mass 325 metabolite of

phenylbutazone actually represents two oxidation products:

one in which one of the aromatic rings has been hydroxylated

(oxyphenbutazone), and another in which the four-carbon chain

was hydroxlated (4-hydroxyphenylbutazone). These are the

only two known metabolites of phenylbutazone (33,34).












Table 6
PARENT IONS OF PHENYLBUTAZONE FRAGMENTS

Fragment
e Mass Parents (abundance, %)

extract 309(100) 337(15)

211(12) 293 (30),309(100),325(15),33

190(38) 309(100),337(38)

188(8) 309(100),325(8),337 (22),349

120(24) 309(100),337(18),349(15)

94(40) 206(55),309(75),325(100),33


7(40)



(12)



7(21),353(42)


Urine Extract


309(100)

211(32)


190(38)

188 (20)

120(100)

94(100)


337(20),349(8)

293 (20),307(88),309(75),325(100),337(19),

349(4)

263(26),309(100),325(9),337 (20),349(5)

234(9),307 (24),309(100),325(23),337(10)

138 (8),177 (18),309(32),325(5),337(4)

138 (48),234(30),309(53),325(40)


Sampl

Serum E








Urine

Parent spectra of the structural subunit masses of

procaine in Table 3 were collected from a urine extract.

The resulting masses of possible metabolites are shown in

Table 5 and Figure 30. Masses 265 and 277 are the (M+29)+

and (M+41)+ ions of procaine and 237+ the MH+ ion. The

peaks from the daughter spectra of the possible metabolites

are listed in Table 7 with those of the MH+ ion. The 235

parent appears to be the product of a dehydrogenation of one

of the terminal ethyl groups attached to the nitrogen. The

daughter spectra of procaine and the mass 235 parent are

shown in Figure 31. The mass 100 peak in the procaine

spectrum has been replaced with a mass 98 peak and the 138

peak is absent. Based on the fragmentation of procaine,

the mass loss must be from one of the two terminal ethyl groups.

This parent is probably a metabolite of procaine since it

appears as a parent ion for several structural subunits and

does not appear in the daughter spectrum of procaine. It

has not been determined, however, if the hydrogen loss is a

direct result of a biochemical process, or the result of a

dehydration of a biochemical oxidation.

Three of the parent masses (38, 196 and 209) represent

metabolites of procaine. Their daughter spectra are shown

in Figure 32. The 138 parent is para-aminobenzoic acid

(PABA), a major metabolite of procaine (35). The mass 196

parent is a carboxylic acid formed following a deamination

of the procaine sidechain. The 209 parent is the product






79























.4)





Ln
u





























o
C a4

















0
f4











*1 -u














SD M0














4-I
:C1
U,




** (

























W w u
e: d d 1
c; rn3
Cn
*H
&<













Table 7
DAUGHTER IONS OF POSSIBLE PROCAINE METABOLITES

Parent Daughter Ions (abundance, %)

235(14) 164(100),120(38),98(4)

224(20) 148(10),120(100),138(6),107(18),90(4),76(80)

221(4) 164(100),120(56),84(6)

209(11) 164(13),138(19),120(100),72(41)

196(46) 138 (32),120(100)

178(16) 120(100)

138 (92) 120(100),94(70),77(14)

237(30)-MH+ 164(42),138(13),120(62),100(100)


















PROCAINE


H2N- -C-O-CHzCH2N(C2H5 2


50.0-


(b)


nVE 50


1I
150


2C0


Figure 31: Daughter spectra of (a) procaine
parent from urine extract.


250


and (b) 235+


(a)


I I L I










103.0- 10
(a)





59.9








W1/E 59 199 159 28 250

(b)





so.e-
1:1




T 93


I/E 50 18 159 2 250
108. E:
(c)










1:8
164


l/E 58 109 150 2C8 258
Figure 32: Daughter spectra of (a) the 138 parent, (b) the
196+ parent and (c) the 209+ parent from urine
extract.








of an N-dealkylation, the removal of an ethyl group from

the procaine sidechain. The appearance of the (M+29)+ ion

for these compounds is a good indication that the parent

ions examined are actually molecular ions.

Two other parents, mass 178 and 221,may be metabolites

of procaine, but their structures have not been determined.

Their daughter spectra are shown in Figure 33. The daughter

spectrum of the 178 parent has only one major fragment of

mass 120, indicating that the procaine sidechain has been

altered. The daughter spectrum of the 221 parent has both

164 and 120 fragments and is very similar to the daughter

spectrum of parent 235. The mass difference corresponds to

a loss of 14, or a methyl group, from the 235 parent, and

total mass loss of 16 from procaine. The similarity of

these compounds to procaine indicates that they are structurally

related, but additional samples must be run, or additional

information on the metabolic process obtained, before a

determination can be made.

Parent spectra of the structural subunit masses of

phenylbutazone in Table 3 were collected from urine extracts.

The resulting masses of possible metabolitesare shown in

Table 6. Masses 337 and 349 are (M+29)+ and (M+41)+ ions

of phenylbutazone. The peaks from the daughter spectra of

the possible metabolites are listed in Table 8 along with

those of the MH ion. The daughter ions of the 325 parent

are a mixture of the ions from two oxidation products of


















(a)


150 1 860


250


Figure 33: Daughter soectra of (a) the 178+ parent and
(b) the 221+ parent from urine extract.


50.0-


W/E 50














DAUGHTER IONS OF

Parent

325(100)



307(58)

293 (100)



263 (82)

234(36)

177 (22)

138 (100)

309(100)-MH+


Table 8
POSSIBLE PHENYLBUTAZONE METABOLITES

Daughter Ions (abundance, %)

307(34),265(16),253 (29),211(19),185(66),

141(18)

265(10),253(100),211(68),188(20)

265(6),211(87),174(85),146(19),132(34),

120(15),83(28)

207(32),190(34),164(48),149(100),100(82)

216(5),188(8),162(8),131(100),94(18),76(19)

160(10),149(100),121(13)

120(58),94(60)

253 (4),216(5),211(19),190(21),188(29),

162(4),120(25)









phenylbutazone which were discussed earlier. The 307

parent is another dehydrogenation product, but the location

of the loss is not clear.

There is no indication from their daughter spectra that

the 293,263,234,177 or 138 parents are structurally related

to phenylbutazone. The 138 parent, however, is PABA which

appears as a parent of the mass 100 structural subunit, and

is a metabolite of procaine.

Parent spectra of the structural subunit masses of diethyl-

carbamazine in Table 3 were collected from urine extracts.

The resulting masses of possible metabolites are shown in

Table 4. Masses 228 and 240 are the (M+29)+ and (M+41)+ ions

of diethylcarbamazine. The peaks from the daughter spectra

of the possible metabolites are listed in Table 9. The mass

186 parent is the same N-dealkylation product detected in

the serum analysis. The 198 parent is another dehydrogenation

product which appears to be the first step in a series of

metabolic steps which break-dj.wn the diethylcarbamazine. An

N-dealkylation occurs, removing either the methyl group

attached to the piperazine ring or one of the ethyl groups

attached to the sidechain. If the methyl group is removed,

producing a metabolite with an MH+ of 184, the metabolite

appears to be eliminated without further reaction, but if

the ethyl group is removed, producing a metabolite with an

MH+ of 170, a second N-dealkylation can occur, removing the

second ethyl group and giving a metabolite with an MH+ of

142. A deamination produces an acid metabolite with an














DAUGHTER IONS OF

Parent

214(55)

212(98)



198(24)

186(12)

184(5)

170(28)

142(58)

141(100)

200(39)-MH+


Table 9
POSSIBLE DIETHYLCARBAMAZINE METABOLITES

Daughter Ions (abundance, %)

141(9),127(4),113(18),100(100),72(18)

183 (19),141(73),127(33),113(18),100(100),

72(25)

127 (38),125(10),100(100),97(8),72(7)

113(14),100(100),72(14)

127(14),113(46),100(100),72(18)

139(57),127 (22),113(100),99(94),70(22)

127 (82),113 (19),100(70),7 2(62)

113(45),99(18),70(45)

127(18),100(100),72(15)









MH of 143 (Figure 34). The mass 212 and 214 parents are

(M+29)+ ions from the 184 and 186 parents. The 141 parent

appears to be structurally related to diethylcarbamazine,

but it is so far unidentified. It is a parent of the mass

113 structural subunit which was selected as such only after

its identification as a fragment of a major metabolite. This

should aid in the identification of any subsequent metabolites

of this structure even if its fragments do not resemble those

of the original drug.

Pharmacokinetics

Pharmacokinetic studies are generally carried out on serum

samples since samples can be collected at regular intervals

and these samples give a true indication of the drug or

metabolite concentration in the body. These studies are used

to determine the amount of time a drug maintains a physiologi-

cally active level in the blood. During this study, the

concentration of three drugs was monitored over a six hour

period after ingestion. Serum samples collected at one hour

intervals were screened using SRM.

Only samples collected one and two hours after ingestion

showed low levels of procaine. After three hours, the

procaine level had decreased to a level below that which could

be detected by the extraction of a 1 mL serum sample. This

indicates that procaine has a short half-life, probably

about one hour, in canines.







Diethylcarbamazine
0 C H
4 /25
C-N
I -C-2


N
I
CH3
199

Metabolites


0 /CzH5
C-N
'CzHs


1
H
183


0
C-N

I


CH3
197


I

CH3
141


,CzHs

SC2H5


C2H


169


0 ,C2H5
C-N



I

H
185


0
-OH
I
0

I

CH3
142


Figure 34: The structures of diethylcarbamazine and its
metabolites.








Diethylcarbamazine levels were monitored and the

concentration plotted versus time. The plot is shown in

Figure 35. It shows that the diethylcarbamazine concentra-

tion reaches a maximum about two hours after ingestion and

then decreases with a half-life of just over two hours.

This is almost identical to values obtained using the current

complex and time consuming methods of analysis (36).

A plot of phenylbutazone concentration vs time is shown

in Figure 36. The phenylbutazone reaches a maximum concen-

tration after about 3.5 hours and has a half-life about 2.5

hours. Again, these values correspond to published values (37).





91









-I






t'-


-4
(U.
*o

4-4
0

0
.H
41
-i

$4



D ,o
0

= a









,I


OU
ON








r-4
Cf





0






0 U
$4
0l 0
O- o
Sr-q*





'IVN91S IS














4-)
c,




0
rl
4-4













Cd

4-)
0 0





*4


S(14
/Cd














0
/c




0 0


W ai








0










4-1
0 u0








(tOIX) 1IVNSIS$
^^ ^ ^ a)















CHAPTER 4
CONCLUSIONS AND FUTURE WORK

The use of triple quadrupole mass spectrometry as a

rapid, sensitive screening technique has been demonstrated

in this study. Screening can be carried out more rapidly

than with current procedures while demonstrating equivalent

or better sensitivity. The screening of urine samples,

which are presently used in post-race screening in Florida,

produces very good results with little or no sample prep-

aration while serum analysis requires an acid-base extraction

and sample concentration to be effective.

Metabolite identification and pharmacokinetic studies

are also promising areas of study for the triple quadrupole

mass spectrometer. Metabolite identification,which currently

may require a month or more,can be carried out in an afternoon,

while pharmacokinetic studies can be carried out rapidly using

the SRM screening procedure developed during this study.

Several changes and additions need to be made to the

current screening procedure. The effect of increasing the

collision energy from 18 eV to 25 eV for some drugs, such

as morphine and theobromine, should be evaluated with the

purpose of decreasing their detection limits since, with








the software currently available, the collision energy for

each drug can be varied during the screening procedure. The

procedure for the automation of the screening procedure

should be completed and the number of standard CAD drug

spectra in the reference library should be increased.

The next step in the evaluation of the screening procedure

should be the acquisition of some samples from one of the

racing tracks. The results from the screening of these

samples should then be compared with the results obtained by

the Florida State Division of Parimutual Wagering Laboratory

in Miami.

Further pharmacokinetic studies should also be carried

out with the cooperation of either the College of Veterinary

Medicine or I.F.A.S. Animal Research. A study of a drug

and its metabolites, showing the disposition of the drug and

each of its metabolites over an extended period of time, i.e.

24 hours, would be very interesting.